Characterizing the Brain with Parametric Mapping of Divergence in Phase-Shift Sensitive Magnetic Resonance Imaging

Phase shift sensitive magnetic resonance imaging, such as multi-direction diffusion imaging, where many directional measurements at varying angles are acquired at each pixel/voxel location, is widely relied upon in research and in the clinic. Analytic methods for parametric mapping, such as scalar image reconstruction, are widely used, producing a variety of image types such as fractional anisotropy, the apparent diffusion coefficient, kurtosis imaging, etc. In this study, we present two novel scalar image reconstruction techniques for parametric mapping of multi-direction phase shift sensitive magnetic resonance imaging examinations, producing phase shift divergence images, and associated sink/source images. The techniques developed were evaluated on a large (n=642) clinical neurological magnetic resonance imaging (MRI) dataset. Results demonstrate potential for novel characterization of water movement properties through these two new image types, which can complement existing techniques. These new images may be sensitive to water flow and microstructural architecture of the brain. While potential utility is demonstrated on a large clinical MRI dataset with statistical analyses, extensive validation is required to confirm utility and validity as part of future work.